Rearranged code

common.go

@@ -1,40 +1,15 @@
 package slinguist
 
-const OtherLanguage = "Other"
-
-var (
-	ExtensionsByLanguage map[string][]string
-	ignoredExtensions    = map[string]bool{
-		".asc": true, ".cgi": true, ".fcgi": true, ".gml": true, ".fx": true,
-		".vhost": true,
-	}
-	auxiliaryLanguages = map[string]bool{
-		"Other": true, "XML": true, "YAML": true, "TOML": true, "INI": true,
-		"JSON": true, "TeX": true, "Public Key": true, "AsciiDoc": true,
-		"AGS Script": true, "VimL": true, "Diff": true, "CMake": true, "fish": true,
-		"Awk": true, "Graphviz (DOT)": true, "Markdown": true, "desktop": true,
-		"XSLT": true, "SQL": true, "RMarkdown": true, "IRC log": true,
-		"reStructuredText": true, "Twig": true, "CSS": true, "Batchfile": true,
-		"Text": true, "HTML+ERB": true, "HTML": true, "Gettext Catalog": true,
-		"Smarty": true, "Raw token data": true,
-	}
+import (
+	"path/filepath"
+	"strings"
 )
 
-func init() {
-	for l, _ := range ignoredExtensions {
-		languagesByExtension[l] = []string{OtherLanguage}
-	}
+// OtherLanguage is used as a zero value when a function can not return a specific language.
+const OtherLanguage = "Other"
 
-	ExtensionsByLanguage = reverseStringListMap(languagesByExtension)
-}
-
-// GetLanguageExtensions returns the different extensions being used by the
-// language.
-func GetLanguageExtensions(language string) []string {
-	return ExtensionsByLanguage[language]
-}
-
-// GetLanguage return the Language for a given filename and file content.
+// GetLanguage applies a sequence of strategies based on the given filename and content
+// to find out the most probably language to return.
 func GetLanguage(filename string, content []byte) string {
 	if lang, safe := GetLanguageByModeline(content); safe {
 		return lang
@@ -60,12 +35,110 @@ func GetLanguage(filename string, content []byte) string {
 	return lang
 }
 
-func reverseStringListMap(i map[string][]string) (o map[string][]string) {
-	o = map[string][]string{}
-	for key, set := range i {
-		for _, value := range set {
-			o[value] = append(o[value], key)
-		}
+// GetLanguageByModeline returns the language of the given content looking for the modeline,
+// and safe to indicate the sureness of returned language.
+func GetLanguageByModeline(content []byte) (lang string, safe bool) {
+	return getLanguageByModeline(content)
+}
+
+// GetLanguageByFilename returns a language based on the given filename, and safe to indicate
+// the sureness of returned language.
+func GetLanguageByFilename(filename string) (lang string, safe bool) {
+	return getLanguageByFilename(filename)
+}
+
+func getLanguageByFilename(filename string) (lang string, safe bool) {
+	lang, safe = languagesByFilename[filename]
+	if lang == "" {
+		lang = OtherLanguage
+	}
+
+	return
+}
+
+// GetLanguageByShebang returns the language of the given content looking for the shebang line,
+// and safe to indicate the sureness of returned language.
+func GetLanguageByShebang(content []byte) (lang string, safe bool) {
+	return getLanguageByShebang(content)
+}
+
+// GetLanguageByExtension returns a language based on the given filename, and safe to indicate
+// the sureness of returned language.
+func GetLanguageByExtension(filename string) (lang string, safe bool) {
+	return getLanguageByExtension(filename)
+}
+
+func getLanguageByExtension(filename string) (lang string, safe bool) {
+	ext := strings.ToLower(filepath.Ext(filename))
+	lang = OtherLanguage
+	langs, ok := languagesByExtension[ext]
+	if !ok {
+		return
+	}
+
+	lang = langs[0]
+	safe = len(langs) == 1
+	return
+}
+
+// GetLanguageByContent returns a language based on the filename and heuristics applies to the content,
+// and safe to indicate the sureness of returned language.
+func GetLanguageByContent(filename string, content []byte) (lang string, safe bool) {
+	return getLanguageByContent(filename, content)
+}
+
+func getLanguageByContent(filename string, content []byte) (lang string, safe bool) {
+	ext := strings.ToLower(filepath.Ext(filename))
+	if fnMatcher, ok := contentMatchers[ext]; ok {
+		lang, safe = fnMatcher(content)
+	} else {
+		lang = OtherLanguage
+	}
+
+	return
+}
+
+// GetLanguageByClassifier takes in a content and a list of candidates, and apply the classifier's Classify method to
+// get the most probably language. If classifier is null then DefaultClassfier will be used.
+func GetLanguageByClassifier(content []byte, candidates []string, classifier Classifier) string {
+	return getLanguageByClassifier(content, candidates, classifier)
+}
+
+// GetLanguageExtensions returns the different extensions being used by the language.
+func GetLanguageExtensions(language string) []string {
+	return extensionsByLanguage[language]
+}
+
+// Type represent language's type. Either data, programming, markup, prose, or unknown.
+type Type int
+
+// Type's values.
+const (
+	Unknown Type = iota
+	Data
+	Programming
+	Markup
+	Prose
+)
+
+// GetLanguageType returns the given language's type.
+func GetLanguageType(language string) (langType Type) {
+	langType, ok := languagesType[language]
+	if !ok {
+		langType = Unknown
+	}
+
+	return langType
+}
+
+// GetLanguageByAlias returns either the language related to the given alias and ok set to true
+// or Otherlanguage and ok set to false if the alias is not recognized.
+func GetLanguageByAlias(alias string) (lang string, ok bool) {
+	a := strings.Split(alias, `,`)[0]
+	a = strings.ToLower(a)
+	lang, ok = languagesByAlias[a]
+	if !ok {
+		lang = OtherLanguage
 	}
 
 	return